Continental Bias of North American and European Dental Research Journals

Introduction: A small pilot study of 2 journals (JDR, EJOS; Y2002) found authors cited references (p<0.0001) from first authors on the same continent more frequently [Leehacharoenkul et al., JDR 2005:Abstr 3120]. A thorough study of several journals and two years is now complete.

Objectives: (1) Determine if research “article-first-authors” (afa) cited “reference-first-authors” (rfa) with differing frequencies within geographic affiliations; and (2) pool journals to detect geographic influences between NA (JDR, JPD, OD, AJD) and EU (EJOS, DM, CR, JD) on these patterns.

Methods: Articles (N=653) and their references (excluding case reports, reviews and non-research items) in 8 journals published during 2 years (1998, 2002) were classified (using ISI and PubMed databases) by first-author geographic affiliation (North America [NA], Europe [EU], Oceania [OC= SE Asia], or other). Each article's contribution (afa) was weighted equally (% rfa).

Results: Results for both years (1998, 2002) were almost identical. Y2002 results are shown with ANOVA for article-first-author sets (rows). NA and EU authors (both afa and rfa) clearly dominated. Journal articles from NA [NA-NA (rfa=1912) vs EU-EU (rfa=986), p<0.01) and EU [NA-NA (rfa=348) vs EU-EU (rfa=1787), p=0.06) cited reference-first-authors from their own continent-of-origin more frequently. Pooled journals for NA versus EU showed geographic trends toward bias.

Article-First-Authors

% Reference-First-Authors, 2002

Pooled NA Journals (afa=402; rfa=8403)

Pooled EU Journals (afa=244; rfa=5598)

NA(rfa)

EU(rfa)

OC(rfa)

Other(rfa)

NA(rfa)

EU(rfa)

OC(rfa)

Other(rfa)

NA(afa)

60±18,a

29±17,b

7±9,c

3±5,c

51±19,a

38±20,b

9±10,c

3±4,c

EU(afa)

40±18,a

50±19,b

7±8,c

2±4,c

31±18,a

59±21,b

8±9,c

2±4,c

OC(afa)

41±16,a

27±15,b

29±17,b

3±5,c

41±17,a

30±18,b

27±17,b

2±3,c

Other(afa)

49±17,a

30±15,b

13±11,c

9±7,c

37±18,a

45±19,b

7±7,c

11±10,c

Discussion: Geographic bias may affect reader impressions of the pertinent literature and distort key citation indexes such as scientific impact factors.

Conclusion: Pertinence and relevance are not the sole determinants of reference selection.

(Support: ISI provided special database access.)